All of my grant-making will go towards reducing existential risks from artificial intelligence. Some reasons to think I'll do a good job at that:

• I have a good deal of grant-making experience, having previously served as a Fund Manager for the EA Long-Term Future Fund (example grant write-ups: https://forum.effectivealtruism.org/posts/HYKDh2mLjapsgj9nB/long-term-future-fund-july-2021-grant-recommendations#Grant_reports_by_Evan_Hubinger and https://forum.effectivealtruism.org/posts/ddBLtdQjcjvZH5JvF/long-term-future-fund-december-2021-grant-recommendations#Grants_evaluated_by_Evan_Hubinger) as well as an FTX Future Fund regrantor.

• I also have a great deal of professional AGI safety research experience, including both very theoretical work (e.g. "Risks from Learned Optimization" (https://arxiv.org/abs/1906.01820) and other stuff I did at MIRI) and very empirical work (e.g. "Discovering Language Model Behaviors" (https://arxiv.org/abs/2212.09251) and other stuff I'm currently doing at Anthropic). As a result, I think I'm particularly well-placed to evaluate AI safety projects across the full spectrum of possible approaches.

• I do a lot of mentorship (e.g. via the SERI MATS program I helped start (https://www.alignmentforum.org/posts/FpokmCnbP3CEZ5h4t/ml-alignment-theory-program-under-evan-hubinger)) and as a result I see a lot of early-stage researchers and projects that would often benefit greatly from additional funding but are too illegible for traditional funders.